HIV-infected adults in the United States corrections system are predominantly active drug users and people of color. These are the very populations with HIV who are not benefiting from effective treatments for HIV, such as highly active antiretroviral therapy (HAART). Jail may be an excellent site for the introduction of medical care for HIV to marginalized populations, particularly drug users who access care for HIV infection at lower rates than other populations of HP/-infected persons. Both primary medical care and initiation or continuation of treatment with HAART may be offered in jaid. The jail setting also provides an ideal opportunity to evaluate the best way to deliver care in order to maximize the benefits both while in jail and, perhaps more importantly, after release from jail. Directly observed therapy (DOT), in which every dose of medication is observed, has been shown to decrease HW viral replication in incarcerated inmates. Other benefits of DOT include sustained HIV viral control that minimizes the likelihood of developing drug resistance to HAART medications started in jail. In the San Francisco City and County Jails, DOT is standard care for inmates on HAART. Unfortunately, our pilot data suggest that the benefits of DOT are often not sustained after inmates are released from jail and must transition to self -administered therapy. Alternatively, a structured program of self-administered therapy in jail may be an equally effective strategy as DOT while inmates are in jail and, may enable inmates to maintain virologic control after they are released from jail. We propose to test the effects of an intervention for delivering HAART to HP/-infected persons in jail (structured self-administered therapy), as compared to usual care (DOT), on virologic and immunologic outcomes in jail and after release from jail. The specific aims of this randomized, controlled trial of HAART in jailed drug users are; A. 1 Primary Aim 1: To compare the effects of structured self-administered therapy as compared to DOT on virologic and immunologic outcomes and incidence of developing new resistant mutations after release from jail. A.2 Primary Aim 2: To compare the effects of structured self-administered therapy as compared to DOT on virologic and immunologic outcomes while subjects are in jail. A.3 Secondary Aim: To measure other factors that may be associated with the short-term and long-term virologic and immunologic outcomes. Such covariates include: demographic factors (including housing and employment); drug and alcohol use; general health status (physical and mental health status); and medications (lifetime and current HAART and medication adherence).